Standing on the Shoulders of Giants: Creativity, Copyright, and AI
How AI Joins the Human Story of Creativity
Human creativity has always been a collaborative and iterative process. From the earliest cave paintings to the most cutting-edge technologies, we have consistently borrowed, adapted, and built upon each other's work. This is not theft—it is the natural evolution of ideas and innovation. Whether it's Shakespeare riffing on classical mythology or Picasso reimagining African art forms, the history of human creativity is a story of shared inspiration.
This shared pool of creativity is both the foundation of and the challenge for copyright law. On one hand, copyright exists to protect individual creators' contributions, incentivizing innovation by granting limited exclusivity. On the other, copyright acknowledges that no creative work is made in a vacuum. It balances protection with the principle of fair use, allowing certain uses of copyrighted material to foster new expression, education, and commentary. Now, AI challenges us to reexamine this balance as it taps into the collective well of human creativity to generate art, ideas, and innovations.
The Creative Borrowing That Built Civilization
To understand AI's role in creativity, we must first accept a simple truth: humans have always borrowed from one another. Consider these examples:
Art: Renaissance artists like Michelangelo studied ancient Greek and Roman sculpture for inspiration. Later, the Impressionists were influenced by Japanese ukiyo-e prints, adopting techniques like flat planes and bold compositions to revolutionize Western painting.
Literature: Shakespeare borrowed liberally from existing works, including Italian novellas and historical chronicles, to create masterpieces like Romeo and Juliet and Macbeth. Even Homer’s Iliad and Odyssey were likely stitched together from earlier oral traditions.
Technology: The Wright brothers built the first airplane using insights from centuries of engineering knowledge, from Da Vinci’s sketches of flying machines to advancements in bicycle mechanics.
Each of these innovators borrowed from the past to create something new, a process AI mirrors in its data-driven generation of art, writing, and ideas.
AI and the Pool of Creativity
At its core, AI is a tool that learns patterns from existing data to generate new outputs. In that sense, it is not fundamentally different from human creators. Just as a writer reads dozens of books before crafting their own novel, AI analyzes vast datasets to produce something novel. The key distinction is scale: while humans rely on memory and subjective experience, AI can process billions of data points in seconds, synthesizing them into something entirely unique.
So while critics argue that AI "steals" from human creators, this misunderstands how creativity works for humans. For example, when a graphic designer is inspired by Bauhaus minimalism or a writer echoes the existential themes of Camus, we don’t accuse them of theft—we recognize their work as part of a broader creative conversation. AI operates similarly, though the conversation it engages in is broader and faster than any human can achieve.
Copyright Law and the Evolution of Creativity
In copyright law, Original Authorship is the threshhold question for granting that right. The work must be created by the author (or multiple authors) without simply copying existing works. If I print a copy of Andy Warhol’s Double Elvis, I can’t get a copyright. If I use it or sell it, I will be infringing the copyright that exists for the original.
But if I paint a woman in the style of Double Elvis, that work would qualify for copyright protection because my creative choices in composition, perspective, color, and detail constitute original authorship. I wouldn’t be infringing Warhol’s copyright because my painting is a transformative work that uses Warhol's piece as inspiration.
When AI generates a painting in the style of Andy Warhol, it does so not by copying any single Andy Warhol work but by analyzing and synthesizing patterns across thousands of examples. Much like my painting, it identifies underlying themes and structures, reimagining them into something distinct—but on a scale and speed far beyond human capability.
This does raise some legal and ethical questions, however:
How do we handle the Human-AI collaboration? In many jurisdictions, copyright requires human authorship meaning AI works may not qualify for protection. (The Constitution of the United States uses the phrase "Authors and Inventors” but doesn’t define whether only a human being can be an author.) I gave the art direction for the image above, does that qualify as a human creating this work?
If I work with other artists to create something new, we all share in the copyright of the final product, even if my contributions were art direction rather than execution. Copyright law recognizes shared contributions and grants protections to the resulting work.
Similarly, when humans work in collaboration with AI to produce creative outputs, the human can and should own the copyright. This approach aligns with existing copyright principles, ensuring that the human creativity guiding the AI remains protected, while fostering innovation through such partnerships.
Does training AI on copyrighted material constitute infringement? Some argue that AI training is akin to fair use, as it transforms raw data into something new. Others see it as exploitation without compensation. Humans can look at paintings, read books, use software and listen to music for inspiration, without infringing any copyrights. How is AI different?
Striking the Balance: AI as an Extension of Human Creativity
Rather than viewing AI as a threat to human creativity, we should see it as an extension of our collective imagination. Just as artists once feared photography would render painting obsolete or novelists worried film would replace books, AI represents a shift—not an end—in how we create.
Copyright law will need to evolve to address these new realities. Policymakers might explore frameworks that compensate creators whose work is used in AI training, much like how music royalties are distributed when songs are sampled. At the same time, we must ensure that these protections do not stifle innovation by locking up the creative commons AI depends on.
Conclusion: Collaboration, Not Competition
AI isn't replacing creativity; it's amplifying it. Like all creators before it, AI relies on the collective achievements of humanity to push boundaries and explore new frontiers. Instead of fearing this technology, we should embrace it as part of the same process that has always driven us forward: standing on the shoulders of giants. As copyright law continues to adapt, we have an opportunity to ensure that this new chapter in creativity honors the past while building an even more innovative future.